thetechnobear's Recent Posts
Ive recently developed stuff for the EigenD/Eigenharp using libusb (including isochronous io), currently Ive tested with linux (including PI2) and works fine.
For the Eigenharps, Eigenlabs wrote (contracted? unfortunately its the only bit I don't have the code for) a small driver, which basically replicated an iso transport into userspace with a 'generic driver'
my understanding though, is that from Windows 8.1 this is no longer necessary, as WinUSB supports it... Im quite tempted to now test this, as libusb* supports this, so theorectically my code should work as is :)
*this assume libusb, has been updated to support the new iso functions in winusb, Ive seen "issues" asking for it to be added, but have not checked the code yet to see if it was.
I think libusb would be the way forward, when this happens as you will get both linux and windows (and also potentially OS X) with the same code base, but it does look like it would be 8.1 would be the minimum supported version, without going the Eigenlabs route.
btw, it may be possible the Eigenlabs driver could be used, as it has no specific eigenharp code* in it, it just forwards requests from a userspace layer into the generic driver...
*hmm we'd have to check its not castrated the vendor ID... you could perhaps even talk to John to see if he might work with you on a derivative driver?
in midi mode , it appears fast slides are generating 2 touches. (18.104.22.168)
( I say midi, its probably unrelated to midi, as you will see below - but its where its easiest to see/hear)
e.g. say you fast slide across the entire surface, you will see (correctly) one slide for the whole slide, but you will see (incorrectly) additional small slides (which look almost independent)
params: quantize on, glissando off, note lock off (everything else default)
replication: easy use 22.214.171.124 midi mode, aalto, patch default, sustain=max, rel=0 and voice = 1 (important) , then slide fast. you will hear the note stop before you reach the end.
ive not fully debugged, I can see in midi monitor 'additional' touches are on different midi channels - this seems to indicate msg->mData is different, as this is used for channel i.e. different touches.
this is also 'confirmed' by the fact if you set max touches = 1, then the problem goes away... presumably as the rogues additional touches are being ignored by the touch tracker.
but oddly only one touch shows on the graph,
not a big problem, but seeing as you are looking at touch tracking, perhaps something to test
it was a fantastic weekend, "we were like kids in a candy store", really inspiring to play so many different instruments.
fascinating to see how the Soundplane/Continuum/Eigenharps all take a different approach to expression, and feel like completely different instruments, each a unique character... we saw that even playing the same synth/patch on different instruments, results in a completely different sound.
Christophe was amazing on the continuum, Ive come away thinking about the techniques he uses, to see if I can perhaps adapt some of them to the soundplane.
oh, and amazing to see how many of us had Aalto/Kaivo paired up with an instrument... there is something about physical modelling and these instruments that just 'feels right'
select carriers - I can't reproduce this one.
hmm, odd, I tried this a few times, and it happened each time. Im now wondering perhaps if it was all 'in the same session' , perhaps a restart of the app would have fixed it, i.e. a quirk, if I get it again, I'll try to get some more info on what is going on.
It's important to me to have nice sensible defaults going forward.
fair enough, more an observation... always tricky, with these kind of things, but as you say better to have it 'correct' going forward.
cool, will give it a go tomorrow
now I can start in earnest on the new algorithm.
played now for a couple of sessions, does feel better...
there appears to be a new bug (126.96.36.199), now if you do "select carriers" ... the SP will initially not do anything, until you also hit 'recalibrate' - not a biggie, but might catch a few out.
cool, will test it :)
Im a bit surprised you have moved the offsets, I know the labelling of the notes was wrong, but I would have thought most, would have got used to note positions, so changing the offset might cause some confusion. (I know you can transpose)
I thought you would probably just change the note names, I know inconsistent with guitar tuning, but id assume its what most users have got used to?
the soundplane software (1.2.4) has introduced a bug which meant it only connects on the default port - Id already reported to Randy. But Ive also now debugged it, and supplied the details of the fix to Randy.
its a trivial bug/fix so hopefully ML will release it
note: you will need to use Aalto 1.6.1, as 1.6.0 also had a bug related to using multiple OSC addresses, and also incorrectly reported 't3d connected, when it wasn't'
ok the issue is in :
void SoundplaneController::doOSCServicesMenu(int result)
the parameters in resolve passed in incorrect order.
Resolve(getServiceName(result - 1).c_str(), kUDPType, kLocalDotDomain);
Resolve(kLocalDotDomain,kUDPType,getServiceName(result - 1).c_str());
have you had a chance to look at the multiple OSC issue I reported via email?
(ie. when using osc offsets, so I can have kaivo and aalto running, worked in previous release of SP software)
if not no issue, then I will fork repo, and fix it... but don't want to duplicate effort.
p.s. any chance of a tag/branch on the madronalib/soundplane repo for released versions. i.e what makes up 1.2.4
love almonds :)
dry by european standards, as no rain from June-August, but some rain the rest of the year.
We run off our own well, which is about 20 meters deep.
btw, today had an issue again... i noticed that the carrier i was uisng (2), had jumped to 0.08, and now carrier 1 was better at 0.04 ... so perhaps these changes are why im getting these issues.
just not why sure the noise changes so much ?
ah, although I'm from the UK, I now live (~2 years) in southern Spain on an old almond farm in the mountains close to the Sierra Nevada - its very beautiful, especially now as the almond trees are in flower.
but... no supplied infrastructure, so I have solar panels (with backup), my own water supply (treatments/storage), and for heating - solar thermal panels, a biomass incinerator which uses almond shells (a 'waste' from the almond farms), and wood burners (to burn cuttings from our trees)... and fortunately, recently, internet via WiFiMax (used to be satellite)
Its alot of fun, but still very much a 'work in progress' :)
hmm, interesting, I can see this behaviour too.
CCs are ignored until the first note comes in...
odd, it looked like an optimisation on Lives part.... but Kaivo doesnt have the same behaviour,
... so either the plugin is telling Live not to send until first note, or aalto is not responding to the CC or flashing the mod light.
@antolyi/randy, just tried here with Mac/1.6.1 vst/Live9 (also N4) ... and works as expected.
I tried first with default MW, ok - then changed CC# to 10, and checked Mod worked on 10, Mod+1 on 11 , both seemed fine.
perhaps a Windows or host (daw) related issue? which daw? does the mod input flash?
Im out of my depth here...so may be talking nonsense :)
UK/US, I assume your referring to mains hum, and the difference between europe using 50hz vs US 60hz... and I guess that implies different carrier frequencies are suitable for europe? Im assuming you have to be careful of harmonics of these frequencies too?
Q. what are the frequencies used by the SP? Q. are the choices of carriers 'hardwired' in the SP, or could we potentially have different choices in europe?
off gird, I was referring to the fact some mains supply cans be pretty 'dirty', with pretty big voltage changes (dips and peaks) as demand alters... and was wondering if this also might induce different RF interferences (and so implicitly change which carrier signals are better)
but as I live in the mountains and don't have mains, I don't have this issue. my solar system has very stable voltage (+/-0.2v variation according to my metrics, which is tiny compared to mains swings which can typically be +/-5v, and at times by much more than that).
Ive read these voltage changes, can make equipment operate differently (sub optimal), but again... Im only going on what Ive read rather than seen any real evidence of it.. not sure if it affects the 'hum' of not.
(the reference to inverters... these, as you possibly know, take my 48v DC to 230v AC, but not all inverters are 'equal', in terms of voltage/cycle variations and some don't even put out perfect sine cycles... so this is something I was very careful of, when designing my system, as well as ensuring I have good earthing, and no ground-loops :))
saying all this, I've not finished building yet, so I'm in a temporary studio, which has a lack of plug sockets, so quite possible the huge number of power strips I have could be far from ideal :)
noise results are:
so carrier 2 is selected.. which I'm 'fairly' sure is always the one selected.
But whats interesting is I'm pretty sure, there used to be 2 (carried 1 and 2) that were around the same 0.004 level, sometimes it would choose one, others time the other.
now 2 things have changed....
a) Ive moved my gear around a bit, due to getting a new mixer in the studio.
b) Ive put a new (powered MTT ) usb hub in place,
I just noticed, if I plug directly into the mac, that carrier 1, goes back down to 0.0048 levels, from 0.00758 ... so I'm wondering if the new hub is creating a bit more 'noise'.
(the usb cable is, I think, the one supplied, and has a ferrite block on it)
so... a couple of questions...
is there anything i can do, to ensure noise is not an issue, or reduce it?
is there anything I should avoid? ( my SP usually is played in from of the mac)
also, next time I see, the issue, I will do a carrier scan, and see what its says, e.g. has the noise gone up.
(when the issue has occurred, Im pretty sure I've not been doing anything to USB devices, or turning any other devices on ... and my power supply is pretty clean, as I'm off-grid running off of solar power with good quality inverters :))
EDIT: actually scrub the 'new hub' theory, just did a few more scans, and even in the new hub its sometimes at 0.004 levels, seems carrier 1 is a bit 'random' on its noise levels.
@mtvic, Live 9.1.7 + Aalto AU 1.6.1 + OSX 10.9.5 works for me :)
sorry not much help ... might be work rescanning/reinstalling?
thanks.... Im pretty sure I was using 1.1.2 before but I guess same place modified url :)
yeah the ghost touches are odd... they occur in a couple of forms...
a) a random touch, quite far away from where other real touches are... i think these tend to be on on the bottom row
b) if play a tri chord, I get 4 or 5 touches (i.e. 2 extra) quite close to the other 3,
(a) is a pain, as the touch tends to persist, until I hit the recalibrate button (pg 2)
also, If i turn up the view scale, it does indeed look like the SP has a touch in that region.
and it also seems commonly in the same place ... bottom row around column 6 (approx)
but it just suddenly starts... I've not placed a touch close to there, and it might start after 5 mins or 30mins!
(b) bit different, but again fine for quite some time of playing, and then suddenly will start glitching, not all the time, but once its starts seems to be more likely to reoccur.
the above are just 'impressions' rather any serious quantifying/testing... but definitely didn't occur in 1.1.2 (which is I'm pretty sure what I was using before)
I did have a try with the reducing template (its about .300) but didn't appear to help, but will try more 'rigoursly' :) threshold Ive got at the default 0.010, but will try to increase.
(I'll look at the ghost touch pressure value a bit close next time, I assume I can use that to tune thresh)
BTW: an idea for you.
I really like the 'pressure value' display on the grid,
I was wondering if we could have something similar for a musical context....
Im trying to get better with quantise OFF, and I'm ok with single touches, as i can tune to a 'backing track'.
BUT when Im doing multiple touches its very difficult, the issue is, I can hear one of the notes is OFF, but its hard to determine which one...
so I was wondering, if we could display a 'tuning' offset for each touch,
this would be a useful training aid, as I could see which note was off, and tune it by ear.
there could be other representations ... e.g. the touch colour could be different for sharp/flat/in-tune.... this would be cool, as very quick to see.
(Im guess not really interested in how much Im out, just which touch and in which direction, just a clue... to then let my ears do the real work :) )
will get back to about the ghost touches, if I can determine anything more concrete to help.
Is there somewhere I can get the previous version from? (I stupidly overwrote it)
Im getting some issues with 'ghost' touches that I didn't get on the previous version, and I need to 'perform' with the soundplane in a couple of weeks, so much as I prefer the new version ... I cannot risk the ghosts touches :(
also, when I go back, will I have to recalibrate?
(It would be nice I think to have a separate app state for each version, in case you need to rollback/ or use different versions for testing)
nice release, already improves touches on the edges for me, and I like the new pressure display. I've emailed you a couple of minor issues, but overall feels really good.
looking forward to seeing and hearing more about the new generation touch tracking :)
my #1 requests would be improved handling of tri-chords, which can be a little hit/miss... but a tough challenge I'm sure, given the 3 notes are quite close.
( though >1 cell apart)
When you log in to the this site, top right you will see a link "my downloads"
(Underneath your login name, my account etc)
apart from this project is going to involve me going down the modular rabbit hole, I'm really excited by it :)
a couple of thoughts
this discussion scares me, Id get a modular to get away from computers, Im not keen on integrating them... am I the odd one out on this?
Id really love it to be simple, and do one job really well, direct connection of the soundplane. Im sure the architecture will be open to expansion anyway, so no need to over complicate at the beginning? keep it simple ... conquer the world later.
In this vain, Id have thought something like:
micro controller with limited storage (need for SP touch tracking software - no?)
upgradeable firmware via USB, and also allow upload of small files (e.g scales/zones), or possibly this could be done via micro SD, but not sure a slot in a non-clean environment is good option, also adds cost.
open source firmware
8 CV out, default assigned to 2x X/Y/Z/Gate BUT by assignable to whatever via firmware (so future firmware could have zones which use as 8 faders)
2 endless encode with small readout to show values, addressable via firmware (example use might be to change scales/zone mapping)
2 CV in (by default, drive the endless encode values)
expansion connector to allow for additional banks of CV outs etc.
Note: Number of CV outs/in I guess this will be driven by cost, and keeping the module 'affordable', which was stated as the goal.
with this kind of model the initial firmware could be very simple (and rock solid!), but could could easily be extended to cover different zone mappings/scales, or other imaginable uses.
open sourcing the firmware, could help relieve some of the pressure on ML.
ML is really creative and is always coming up with fantastic ideas and products, so Im sure we (users) have to be realistic about how much effort can be put into this project (without letting other ML projects suffer)
awesome stuff... OSC and automation working great here... thanks!
ok uploaded here Cantor Digitalis Max Patch
this actually combined with changes also made for a haken continuum
(don't worry there is no mention of SP, as it needs no UI, it just works )
to use, download CD from their website, then copy this patch into the _MainPrograms directly and start it from Max.
you will need the CNMAT packages for OSC-route, since i use those (would be easy to avoid this, but I often use them so didnt think not to this time around :))
then run the SP app in OSC mode, and use the chromatic zone, without quantisation, (as cantor digitalis has its own.)
first 6 columns have vowel space, rest is playing area.
oh cantor digitalis set manual tuning to -14 -3, no link... this will ensure correct tuning for voices.
Ive 'ported' this to the Soundplane, as its really fun, you can get some amazing voice sounds, with vowel control on one x/y grid, and pitch/effort and larynx on another.
go check out their website (in YT description) for some examples of what skilled players can do with it, very impressive stuff.
If your interested I could post the patch…
(Im not sure if max patches can be posted here, as the forum does some weird formatting)
I use OSC, put the soundplane into chromatic zone (so I get full x/y/z), then Ive edited the main control patch to send the relevant max messages based on the tch messages. If your familiar with max its simple enough.
as I say, though I can post the patch if your interested/prefer
any news on a 1.6.1, which fixes the OSC and automation?
@jue, interesting its working for me with N4Pro (b149), osx 10.9.5 , i.e. I can see aalto and load it within N4, and also save it as part of a stack etc.
though it doesnt appear to want to load the presets from the N4 folders, I guess as these are the .aupresets - this is why I asked Randy, if perhaps I can delete that folder, so it doesnt confuse.
Is automation working?
Ive tried in both Live 9.1.7/Mac 10.9 (au and vst) and Bitwig 1.1, and it doesnt seem to be working for me.
I think I'm doing everything correctly, as it works with Kaivo (and others) and also if I reinstall Aalto 1.5 it works there too.
(in Live the curve goes dark red, but i don't know what that means, I guess its having a problem)
note: the automation initially appears to work, i.e. curve gets drawn, dial moves, but as soon as it loops/playbacks it turns dark red and doesn't do anything. I have the same issue if I just draw the automation in.
edit: works ok in Logic 10.1
ooooohhh... Ive been avoiding modulars, but this might tip me over the edge :)
id echo @drart comments, id really love a box to allow my SP to be computer free, perhaps a separate CV + midi box might widen the appear to other analog (and non analog) synths.
Ive found if you prefix your bank name with Aalto or Kaivo, it considers it to be a factory bank, if its anything else its not.
@Randy, can I now delete ~/Library/Audio/Presets/Madrona Labs ?
i.e its not used by Kaivo or Aalto now?